The PRIsm MUlti-object Survey (PRIMUS)

Transcription

1 The PRIsm MUlti-object Survey (PRIMUS) Alison Coil University of Arizona Steward Observatory March 2008

2 Overview: Galaxy evolution to z ~ 1 is still cosmic variance limited: DEEP2, VVDS, COMBO-17, COSMOS spec-z/photo-z samples aren t enough sq. degrees each For many questions, 1% redshifts are good enough : even clustering statistics are projected over these scales A low dispersion prism can get this accuracy: and allows you to multiplex in the dispersion direction! We ve done a huge survey in this mode with a prism on IMACS: PRIMUS

3 Galaxies have evolved since z=1: Fewer red galaxies at z=1 Galaxies are less clustered at z=1 # density of red galaxies Faber et al redshift Coil et al The dominant error on galaxy evolution since z=1 is cosmic variance!

4 The largest redshift survey of the distant universe is DEEP2: z= , covers 3 sq. degrees of sky and used DEIMOS for 80 Keck nights to obtain 40k redshifts. AGES: Largest intermediate redshift survey - z= Used ~30 nights on the MMT + Hectospec to obtain 25k redshifts. z=0 z=1.5 At intermediate redshifts, you need to survey uncomfortably large areas with 8m class telescopes - unfeasible w/ current instrumentation. At low-redshift, SDSS is the largest redshift survey to z~0.2. Has spectra for 700k galaxies. Provides a detailed picture of the galaxy population today.

5 Why hasn t this been done already? Scaling from AGES - 25,000 galaxies in 30 nights with Hectospec: Should we build another instrument? loads of $$$$$ and time 300,000 galaxies to I=23 would require about ~900 dark nights

6 The volume probed by deep photometry overwhelms that of spectroscopy COSMOS 2 sq. degrees Many square degrees covered by deep multiwavelength observations NDWFS: 9 sq. degrees Spitzer Chandra HST XMM-Newton GALEX

7 SWIRE: Spitzer Wide-area Infra Red Extragalactic Survey SWIRE is a major Spitzer legacy survey: imaged a 50 sq. degrees of sky in 6 fields (3 north, 3 south) to a depth of L* at z~2. Took a month total of Spitzer time. Main goal is galaxy evolution to z>1. But their fields lack redshifts!! The idea was to use photometric redshifts...

8 Photometric redshifts have seen great success Allow redshift estimates for very faint objects and large samples where spectroscopy is observationally too expensive Typical errors of 3-5% are too large to separate populations into all but the most extreme environmental bins and outliers cause systematic errors Brodwin et al 2006

9 1% redshifts are good enough for LSS Measurements of large-scale structure (clustering, local density, environment) integrate over 1% in redshift because of redshift-space distortions. Coil et al. 2008

10 Photometric redshift surveys are useful e.g., COMBO-17 achieves 3% accuracy in redshift w/ 17 bands (broad+narrow bands)

11 But a slitmask + prism is more efficient and provides more accurate redshifts!

12 But a slitmask + prism is more efficient ~150 pixels We achieve 1% redshifts w/ these low-dispersion prism spectra from IMACS on Magellan

13 PRIMUS Goals: Obtain redshifts for an SDSS-like sample of galaxies to z=1 300,000 galaxies over 15 deg 2 Target fields with existing optical imaging Target fields with multi-wavelength data to maximize the legacy value of these datasets (hit all southern SWIRE fields with good optical imaging) PRIMates: Arizona: Alison Coil, Richard Cool, Daniel Eisenstein (PI), Ken Wong MIT: Scott Burles NYU: Michael Blanton, Guangtun Zhu Hawaii: Adam Bolton 39 dark nights from Jan Jan 2008

14 z=1.5 DEEP2 PRIMUS AGES PRIMUS will survey 1/2 the volume of SDSS at 0.2<z<1.0 SDSS z=0

15 Context with other surveys PRIMUS Current high-redshift surveys have over 10x fewer redshifts and volumes than SDSS PRIMUS will rectify this difference 5x larger than DEEP2

16 We have designed and commissioned a new prism for IMACS on Magellan Multiplexing in the spectral direction allows for simultaneous observations of 3000 galaxies Doing a flux-limited survey (i~23) - no color selection.

17 Example Spectra red + blue - 2 N+S obs.

18 CFHTLS D4 SDSS Southern Fields not to scale! DLS COSMOS DEEP2 XMM-Swire DEEP2 ELAIS S1 CDFS

19 Redshift Precision Work in progress!! Using high-resolution redshifts from DEEP2 and VVDS to test our z s Still working on known systematics. This will tighten - haven t corrected for alignment offsets, not using broad-band photometry yet, investigating outliers, etc. Currently have 6% catastrophic outliers and 1% Δz/(1+z) - will improve! Includes both blue and red galaxies.

20 Current Status: Observing for 2 years - finished this January >300,000 spectra over 13 sq. degrees Most of the survey has Spitzer imaging Much has X-ray and GALEX imaging as well Now finalizing reductions and redshifts Final Survey: ~1/2 the volume of SDSS 5x larger than DEEP2 over 10x faster than Keck or VLT! By far the largest faint galaxy redshift survey completed and will enable science with the largest sample of Spitzer redshifts!

21 Just a bit of the PRIMUS science to come: Evolution of multivariate functions of galaxy properties: color, luminosity, stellar mass (Spitzer), star formation rate (GALEX + Spitzer) from z~ Clustering as function of luminosity and/or stellar mass for red and blue galaxies separately LIRG/ULIRG evolution: clustering and stellar mass AGN selected by optical, IR or X-ray emission: comparison of different selections and AGN vs galaxy clustering of various types...stay tuned!